The following documents are incorporated by reference:    [SPVC-IW] Swallow draft “Soft Permanent Virtual Circuit Interworking between PWE3 and ATM”, draft-swallow-pwe3-spvc-iw-00.txt    [PWE3-CONTROL] Martini, L., et al., “Pseudowire Setup and Maintenance using LDP”, draft-ietf-pwe3-control-protocol-05.txt, December 2003    [L2VPN-SIG] Rosen, E., Radoaca, V., “Provisioning Models and Endpoint Identifiers in L2VPN Signaling”, draft-ietf-12vpn-signaling-00.txt    [VPN-AUTO-DISCOVERY] Ould-Brahim, H., Rosen, E., Rekhter, Y., “Using BGP as an Auto-Discovery Mechanism for Provider-Provisioned VPNs”, draft-ietf-13vpn-bgpvpn-auto-00.txt.    [BGP-L2VPN-AD] Radoaca, Unbehagen, P., et al., “BGP-based Auto-Discovery for L2VPNs”, work in progress, draft-hlmu-12vpn-bgp-discovery-00.txt    [SM-SCOPE-REQ] “Service Mediation scope and requirements”, mpls2004.010.00    [SHAH-QOS] Shah, H., Ould-Brahim, H., Metz, C., “QoS Signaling for PWE3”, draft-shah-pwe3-pw-qos-signaling-00.txt, work in progress.
Layer-2 protocols such as Frame Relay and Asynchronous Transmit Mode (“ATM”) were developed with the intent of supporting voice and data communications in service provider networks. Consequently, Layer-2 legacy equipment is widely deployed in existing networks. More recently, service providers have been deploying Internet Protocol (“IP”) and Multi-Protocol Label Switching (“MPLS”) based equipment. Since it would be quite costly to deploy parallel networks, service providers often deploy IP and MPLS based equipment in the same network with legacy equipment based on Layer-2 protocols.
One provisioning model for a network that includes both ATM/FR equipment and IP/MPLS equipment is the unmapped mode. The unmapped mode requires that calls originating from the Layer-2 domain encode information that identifies the target Forwarder and the destination MPE in the MPLS domain. The Forwarder located on the MPE side is identified using Layer-2 related information such as port number, VPI, VCI, and DLCI values. Forwarder identifiers at the MPE and the IP address of the MPE are taken exclusively from the Layer-2 information carried within the native Layer-2 call. The attachment identifiers on the MPE represent information relevant to the Layer-2 network being mediated. The native Layer-2 address represented by the Calling Party Number (in this case the ATM NSAP address) encodes the IP address of the destination MPE and is assigned a specific address format code that indicates that the address contains an IP address. During the signaling phase the mediation function screens the Called Party Information Elements and extract the IP address, the port number and the VPI.VCI values. The MME screens the AFI and ICP from the received call, extracts the IP address representing the loopback address of the destination MPE, and establishes a pseudowire to that MPE. A TAII is constructed from the information carried within the NSAP address (in this case the ESI—End System Identifier) and the SPVC IE (VPI.VCI/DLCI values). An Example of the unmapped mode solution is described in [SPVC-IW], which is limited to ATM technology. However, the unmapped mode has some disadvantages.
The unmapped mode disadvantageously restricts the set of Forwarders on the MPE devices to only those that are relevant to the Layer-2 network. Indeed, it is not possible for the MPLS operator to configure two separate Forwarders for the same mediated service on two different MPEs with the same or different VPI.VCI/DLCI values or port numbers. Neither is it possible for the unmapped mode to support scenarios where an Ethernet port is backing up a primary ATM port on the MPE, or an ATM port is upgraded to an Ethernet port without reconfiguring all the Layer-2 connections on the Layer-2 network domain destined to that MPE. Further, it is not clear how the unmapped mode supports other addressing plans such as E.164 and X.121 which can still be found in some Frame Relay networks today (since this mode requires encoding an IP address of the MPE onto the native Layer-2 address information). Finally, the unmapped mode does not offer the ability to perform address mobility within the MPLS network such as relocating the attachment identifiers to different ports, or to different MPEs, or even to different MPLS networks without requiring modifications to these identifiers.